Chemical Thermodynamics II

1. In this lecture, we shall continue to analyze our chemical thermodynamics bond graphs, making use of bond-graphic knowledge that we hadn’t exploited so far. This shall lead us to a more general bond-graphic description of chemical reaction systems that is less dependent on the operating conditions. The RF-element and the CF-element are explained in their full complexity. Chemical Thermodynamics II

2. Structural analysis of chemical reaction bond graph The chemical resistive field Multi-port gyrators The chemical capacitive field Isochoric vs. isobaric operating conditions Equation of state Adiabatic vs. isothermal operating conditions Caloric equation of state Enthalpy of formation Tabulation of chemical data Heat capacity of air Table of Contents

3.  mix reac MTF CF RF  N mix reac   Is the RF-element truly reactive? A Structural Analysis of theGeneric Chemical Reaction Bond Graphs • Let us look once more at the generic chemical reaction bond graph:

4. C I e = R( f ) q = C( e ) p = I( f ) Resistor: Capacity: Inductivity:   e f q p R Relations Between the Base Variables • Let us recall a slide from an early class on bond graphs: A reactive element must be describable purely by a (possibly non-linear) static relationship between efforts and flows.

5. · p · qi = T ·Si + m · ni 1. Gibbs equation: p, T are e-variables Vi, ni are q-variables. p · Vi = ni· R · T 2. Equation of state: The RF-Element I • Let us analyze the three equations that make up the RF-element: The Gibbs equation is certainly a static equation relating only efforts and flows to each other. It generalizes the “S” of the RS-element. The equation of state is a static equation relating efforts with generalized positions. Thus, it clearly belongs to the CF-element!

6. By differentiating the equation of state: we were able to come up with a structurally appropriate equation: Yet, the approach is dubious. The physics behind the equation of state points to the CF-field, and this is where it should be used. p, T are e-variables qi, ni are f-variables. The RF-Element II p · qi = ni· R · T

7. This also makes physical sense. The equation of state describes a property of a substance. The CF-field should contain a complete description of all chemical properties of the substance stored in it. The RF-field, on the other hand, only describes the transport of substances. A pipe really doesn’t care what flows through it! The RF-field should be restricted to describing continuity equations. The mass continuity is described by the reaction rate equations. The energy continuity is described by the Gibbs equation. What is missing is the volume continuity. The RF-Element III

8. We know that mass always carries its volume along. Thus: Using the volume continuity equation, we obtain exactly the same results as using the differentiated equation of state, since the equation of state teaches us that: thus: · qi = (V/M) · M = (V/n) ·ni  V/n = R · T / p p · Vi = ni· R · T R · T p qi = ni· The RF-Element IV which is exactly the equation that we had used before.

9. What have we gained, if anything? The differentiated equation of state had been derived under the assumption of isobaric and isothermal operating conditions. The volume continuity equation does not make any such assumption. It is valid not only for all operating conditions, but also for all substances, i.e., it does not make the assumption of an ideal gas reaction. mass continuity energy continuity The RF-Element V  nreac= k .* n ; qreac/V= nreac /n; p * qreac = T * Sreac+ mreac .* nreac ; are the set of equations describing thegeneric RF-field, where V is the total reaction volume, and n is the total reaction mass. volume continuity ·

10. The RF-Element VI nreac= k .* n 3. Reaction rate equations: The reaction rate equations relate flows (f-variables) to generalized positions (q-variables). However, the generalized positions are themselves statically related to efforts (e-variables) in the CF-element. Hence these equations are indeed reactive as they were expected to be. Thus, we now have convinced ourselves that we can write all equations of the RF-element as: f = g(e). In the case of the hydrogen-bromine reaction, there will be 15 equations in 15 unknowns, 3 equations for the three flows of each one of five separate reactions.

11. We still need to ask ourselves, whether these 15 equations are irreversible, i.e., resistive, or reversible, i.e., gyrative. We already know that the C-matrix describing a linear capacitive field is always symmetric. Since that matrix describes the network topology, the same obviously holds true for the R-matrix (or G-matrix) describing a linear resistive field (or linear conductive field). These matrices always have to be symmetric. The Linear Resistive Field

12. Let us now look at a multi-port gyrator. In accordance with the regular gyrator, its equations are defined as: e e 1 2 MGY f f R 1 2  e1’ = f2’ · R’  e1’ · f1 = f2’ · R’ · f1= f2’· e2  e2 = R’ · f1 The Multi-port Gyrator I e1 = R · f2 e1’· f1 = e2’· f2 = f2’· e2

13. In order to compare this element with the resistive field, it is useful to have all bonds point at the element, thus: with the equations: e e 1 2 MGY f f R 1 2 The Multi-port Gyrator II f1 = G’ · e2 f2 = -G · e1 e1 = -R · f2 e2 = R’ · f1 or: where: G = R-1

14. In a matrix-vector form: f1 = 0 G’ e1 f2 = -G 0 e2 · skew-symmetric matrix • Any matrix can be decomposed into a symmetric part and a skew-symmetric part: Ms = (M + M’) /2 Mas = (M - M’) /2 where: M = Ms + Mas The Multi-port Gyrator III f1 = G’ · e2 f2 = -G · e1 

15. Example: M’ = 1 2 3 4 1 2 3 4 1 3 2 4 M = 1 2.5 2.5 4 0 -0.5 0.5 0 Mas = (M - M’) /2 = Ms = (M + M’) /2 = is symmetric: (Ms = Ms’) is skew-symmetric: (Mas = -Mas’) 0 -0.5 0.5 0 1 2.5 2.5 4 M = = + = Ms + Mas Symmetric and Skew-symmetric Matrices  

16. Hence given the equations of the RF-element: these equations can be written as: Conductive part • Thus: f = Gs (e) · e + Gas (e) · e Gyrative part The RF-Element VII f = g(e) f = G(e) · e

17. Example: f1 = e12 + 2e2 f2 = -e1 + e22 f1 e1 2 e1 f2-1 e2 e2 · = e1 2 -1 e2 e1-1 2 e2  G(e) = G’(e) = 0 1.5 -1.5 0 e1 0.5 0.5 e2 G(e) = = + = Gs(e) + Gas(e) Conduction matrix Gyration matrix e1 2 -1 e2 The RF-Element VIII 

18. We should also look at the CF-elements. Of course, these elements are substance-specific, yet they can be constructed using general principles. We need to come up with equations for the three potentials (efforts): T, p, and g. These are functions of the states (generalized positions): S, V, and M. We also need to come up with initial conditions for the three state variables: S0, V0, and M0. The CF-Element I

19. The reaction mass is usually given, i.e., we know up front, how much reactants of each kind are available. This determines M0 for each of the species, and therefore n0. It also provides the total reaction mass M, and therefore n. In a batch reaction, the reaction mass remains constant, whereas in a continuous reaction, new reaction mass is constantly added, and an equal amount of product mass is constantly removed. Modeling continuous reactions with bond graphs is easy, since the chemical reaction bond graph can be naturally interfaced with a convective flow bond graph. The CF-Element II

20. Chemical reactions usually take place either inside a closed container, in which case the total reaction volume is constant, or in an open container, in which case the reaction pressure is constant, namely the pressure of the environment. Hence either volume or pressure can be provided from the outside. We call the case where the volume is kept constant the isochoric operating condition, whereas the case where the pressure is kept constant, is called the isobaric operating condition. Isochoric vs. Isobaric Operating Conditions

21. The equation of state can be used to compute the other of the two volume-related variables, given the reaction mass and the temperature: The Equation of State Isobaric conditions (p=constant): p ·V0 = n0· R · T0 Isochoric conditions (V=constant): p(t) · V = n(t)· R · T(t)

22. We can perform a chemical reaction under conditions of thermal insulation, i.e., no heat is either added or subtracted. This operating condition is called the adiabatic operating condition. Alternatively, we may use a controller to add or subtract just the right amount of heat to keep the reaction temperature constant. This operating condition is called the isothermal operating condition. Adiabatic vs. Isothermal Operating Conditions

23. We need an equation that relates temperature and entropy to each other. In general: T = f(S,V). To this end, we make use of the so-called caloric equation of state: where: The Caloric Equation of State I ds = (cp/T) · dT – (dv/dT)p· dp ds = change in specific entropy cp = specific heat capacity at constant pressure dT = change in temperature (dv/dT)p = gradient of specific volume with respect to temperature at constant pressure dp = change in pressure

24. Under isobaric conditions (dp = 0), the caloric equation of state simplifies to: or: which corresponds exactly to the heat capacitor used in the past. The Caloric Equation of State II ds = (cp/T) · dT  ds/dT = cp/T Ds = cp ·ln(T/T0 )  DS = g·ln(T/T0 )

25. In the general case, the caloric equation of state can also be written as: In the case of an ideal gas reaction: Thus: · · · s = (cp/T) · T – (dv/dT)p· p · · · s = cp · (T/T) – R · (p/p)  s - s0 = cp ·ln(T/T0 ) – R ·ln(p/p0 ) The Caloric Equation of State III (dv/dT)p= R/p

26. The initial temperature, T0, is usually given. The initial entropy, S0 , can be computed as S0 = M0· s(T0 ,p0 ) using a table lookup function. In the case of adiabatic operating conditions, the change in entropy flow can be used to determine the new temperature value. To this end, it may be convenient to modify the caloric equation of state such that the change in pressure is expressed as an equivalent change in volume. In the case of isothermal conditions, the approach is essentially the same. The resulting temperature change, DT, is computed, from which it is then possible to obtain the external heat flow, Q = DT · S, needed to prevent a change in temperature. The Caloric Equation of State IV · ·

27. Finally, we need to compute the Gibbs potential, g. It represents the energy stored in the substance, i.e., the energy needed in the process of making the substance. In the chemical engineering literature, the enthalpy of formation, h, is usually tabulated, in place of the Gibbs free energy, g. Once h has been obtained, g can be computed easily: The Enthalpy of Formation g = h(T,p) – T · s

28. We can find the chemical data of most substances on the web, e.g. at: http://webbook.nist.gov/chemistry/form-ser.html. Searching e.g. for the substance HBr, we find at the address: http://webbook.nist.gov/cgi/cbook.cgi?ID=C10035106&Units=SI&Mask=1 Tabulation of Chemical Data I

29. Tabulation of Chemical Data II

30. Equation of state Caloric equation of state Gibbs energy of formation The Heat Capacity of Air I We are now able to understand the CFAir model:

31. The Heat Capacity of Air II  p = T·R·M/V p·V = T·R·M T = T0·exp((s–s0- R·(ln(v)-ln(v0 )))/cv)  T/T0 = exp((s–s0- R·(ln(v/v0 )))/cv)  ln(T/T0 ) = (s–s0- R·ln(v/v0 ))/cv  cv·ln(T/T0 ) = s–s0- R·ln(v/v0 )

32. The Heat Capacity of Air III g = T·(cp – s)  h = cp·T g = h - T·s for ideal gases

33. Cellier, F.E. (1991), Continuous System Modeling, Springer-Verlag, New York, Chapter 9. Greifeneder, J. (2001), Modellierung thermodynamischer Phänomene mittels Bondgraphen, Diplomarbeit, University of Stuttgart, Germany. Cellier, F.E. and J. Greifeneder (2009), “Modeling Chemical Reactions in Modelica By Use of Chemo-bonds,” Proc. 7th International Modelica Conference, Como, Italy, pp. 142-150. References